Aminoglycoside antibiotics. 3. Synthesis of a furanosyl isomer of

Aminoglycoside antibiotics. 3. Synthesis of a furanosyl isomer of kanamycin B from a protected 3-amino-3-deoxyglucofuranosyl chloride. Robert D. Sitri...
3 downloads 0 Views 670KB Size
3048 J . Org. Chem., Vol. 43, No. 15, 1978

Sitrin, Cooper, and Weisbach

Registry No.-2,15981-92-7; 3a, 66323-40-8; 3b,30516-87-1; 4a, 66323-41-9; 4b, 66323-42-0; 5a, 66323-43-1; 5b,66323-44-2; 6,740325-0; 7a,66323-45-3; 7b,66323-46-4;8a,66323-47-5;8b, 66323-48-6; 9, 66323-49-7; 10, 66323-50-0; 11, 52854-12-3; 12, 21047-87-0; 13, 9-(3-Acetylamino-~2,3-dideoxy-~-~-ribofuranosyl)adenine4105-38-8; adenosine, 58-61-7. (10). 9-(3-Amino-2,3-dideoxy-@-~-ribofuranosyl)adenine (6) (35 mg, References a n d Notes 0.14 mmol) was dissolved in methanol (10 mL) and treated with acetic anhydride (0.03 mL, 0.3 mmol) and triethylamine (0.05 mL). The L. Thelander, B. Larson,J. Hobbs, and F. Eckstein, J. Biol. Cbem., 251, 1398 solution was stirred over night at room temperature and evaporated. (1976). The residue was dissolved in a minimal amount of 9% methanolL. Skoog, G. Bjursell, L. Thelander, T. bgerstrom. J. Hobbs. and F. Eckstein, Eur. J. Biocbem., 72, 371 (1977). chloroform and applied to a column of silica gel (15 g) made up in W. W. Lee, A. Benitez, C. D. Anderson, L. Goodman, and B. R. Baker, J. chloroform and was eluted with 200 mL each of 9 and 15% niethaAm. Chem. SOC.,83, 1906 (1961). nol-chloroform, collecting 17.4-mL fractions. Fractions 17-21 conJ. B. Hobbs and F. Eckstein. J. Org. Cbem., 42,714 (1977). tained a material which gave a single spot ( R f0.41 in system B). These R. P. Glinski. M. Sami Khan, R. L. Kalmas, and M. B. Sporn, J. Org. Cbem., fractions were combined and evaporated to dryness. 9-(3-Acetylam38, 4299 (1973). ino-2,3-dideoxy-p-D-ribofuranosyl)adenine (10) was crystallized as M. Miyaki, A. Saito, and B. Shimizu, Cbem Pharm. Bull., 18, 2459 (1970). fine plates from methanol (33 mg, 81%):mp 235-237 "C; UV, , ,A F. W. Lichtenthaler and K . Kitahara, Angew. Chem., 87, 839 (1975). (HzO) 259.5 ( e 15300) and (pH 1) 257 nm ( f 14800);CD , , ,A 265 ([e] T. Azuma, K. Isono, P. F. Crain, and J. A. McCloskey, Tetrahedron Lett., -goo), 226 ([e]+ l l O O ) , and 217 nm ( [ O ] -600); NMR (MenSO-ds) 6 1687 (1976). 1.89 (s,3 H, MeCO), 2.:!-2.9 (m, 2 H, H-Z'a,b), 3.64 (m, 2 H, H-5'a,b), T. Azuma, K. Isono, P. F. Crain, and J. A. McCioskey, J. Cbem. SOC.,Chem. Commun,, 159 (1977). 3.93 (m, 1H , H-4'),4.50 (m, 1H, H-3'),5.17 (brs, 1H, OH-5'), 6.39 (t, T. Azuma, K. Isono, and J. A. McCioskey. Nucleic Acids Res., Spec. Pub/.. 1 H, J1,,pa = J13,~'b= 6.5 Hz, H-l'), 7.31 (brs, 2 H, NH2),8.16,8.38 (s, 19 (1976). 2 H, H-2 and H-8),8.41 (d, 1 H, Js,," = 7 Hz, "-3'). H. C.Marsmann and H.-G. Horn, 2.Naturforscb.,6, 27, 1448 (1972). Anal. Calcd for ClzH 16NS03:C, 49.31; H, 5.52; N, 28.75. Found: C, G. Kowollik, K. Gaertner. and P. Langen, Tetrahedron Lett., 3863 49.15; H , 5.56; N, 28.66. (1969). G. Roesier, Dissertation,Technische Universitat Braunschweig, 1974. Synthesis of Adenosine from 2',3',5'-Tri-O-acetyluridinr:(13). Y. Furukawa and M. Honjo, Cbem. Pbarm. Bull.. 16, 1076 (1968). NWctanoyladenine ( 11) (238 mg, 0.91 mmol) and 2',3',5'-tri-OR. K. Ness in "Synthetic Procedures in Nucleic Acid Chemistry", Voi. 1 , acetyluridine (13)(185 mg, 0.50 mmol) were suspended in acetonitrile W. W, Zorbach and R. S. Tipson, Ed., Wiley, New York, N.Y., 1968. p ( 3 mL) and BSA (0.5 :nL, 2.0 mmol) was added. The mixture was 183. heated at reflux temperature for 15 min. Trimethylsilyl trifluoroG. Schramm, G. Lungmann, and F. Bechmann, Biocbim. Biopbys. Acta, 145,221 (1967). methanesulfonate (0.1L mL, 0.65 mmol) was added to the clear soluW. S. Mungall, G. L. Greene, G. A. Heavener, and R. L Letsinger. J. Org. tion. After heating at reflux temperature for 4 h, the reaction mixture Cbem., 40, 1659 (1975). was poured in 25 mL of ethanol-concentrated NHlOH (4:l) with M. J. Robins and R. K. Robins, J. Am. Cbem. SOC., 87, 4934 (1965). stirring. After 1 day at room temperature, the solution was evaporated L. B. Townsend in "Synthetic Procedures in Nucleic Acid Chemistry", Voi. and the residue was diiisolved in 40 mL of 60% methanol-water and 2,W. W. Zorbach and R. S. Tipson, Ed., Wiley, New York, N.Y., 1973,p 338 ---. applied tu a column of !Dowex 1 X 4 (OH-) (20 mL) which was eluted (20)T. R. Emerson, R. J. Swan, and T. L. V. Ulbricht. Biocbem. Biophys. Res. with 6090 methanol--water (300 mL) and 75% methanol-water (150 Commun., 22, 505 (1966). mL). The main fractions containing adenosine were combined and (21) D. W. Miles, L. B. Townsend, M. J. Robins, W. H. Inskeep, and H. Eyring. evaporated to dryness. The residue was recrystallized from water (75 J. Am. Cbem. SOC.,93,1600 (1971). mg. 56%),mp 236-238 This compound was found to be identical (22)W. C.Brunner and M. F. Maestre, Biopolymers, 14, 555 (1975). (23)A Synthesis of an adenosine derivative through a transglycosylation from in all respects with an ;authentic sample. 2',3',5'-tri-Oacetyl-~-benzoyluridine in a yield of 58% has been reported earlier. See ref 10. Acknowledgment. We thank the Deutsche Forschungs(24)C. A. Dekker, J. Am. Chem. SOC., 87, 4027 (1965). gemeinschaft for financial support and one of us (M.I.) thanks (25)B. R. Baker, Cbem. Biol. Purines, Ciba found. Symp., 120 (1957). the Alexander von Humboldt Stiftung for a fellowship. The (26)This spot was probably the 01 anomer (4a) of the starting material (4b) which was isolated in the synthesis of the guanine derivatives. measurement, of the NMR spectra by Mr. B. Seeger is grate(27)This was confirmed by the alkaline hydrolysis of ethanolic extracts of these fully acknowledged. We also thank Dr. V. Armstrong for spots. (28)All atrempts to separate Ea and 8b were unsuccessful. helping us with the writing of the manuscript. and H-S'a,b), 6.20 (t, Jl,,z,a= J1',2'b = 6.5 Hz, 1H, H-l'),7.88 ( 3 , 1 H, H-8). Anal. Calcd for C1&:1&03: C, 45.11; H, 5.30; N, 31.56. Found: C, 45.24; H, 5.42; N, 31.19.

Aininoglycoside Antibiotics. 3.' Synthesis of a Furanosyl Isomer of Kanamycin B from a Protected 3-Amino-3-deoxyglucofuranosyl Chloride Robert D. S h i n , * David J. Cooper, and Jerry A. Weisbach Research and Development Division, S m i t h Kline & French Laboratories, Philadelphia, Pennsylvania 19101 Received Nouember 21, 1977

A new crystalline glycosylating agent, comprised of a mixture of CY- and P-3-acetamido-2,5,6-tri-O-henzyl-3deoxy-D-glucofuranosyl chlorides (7a,b),was synthesized. It was used to prepare, via a Koenigs-Knorr type conand P-u-gludensation, the 2-deoxy-4-0-(2,6-diamino-2,6-did~~oxy-a-~-glucopyranosyl)-6-0-(3-amino-3-deoxy-cucofuranosy1)-D-streptamines, 1 and 2, isomers of the antibiotic kanamycin B. The structure of a-glycoside 1 was confirmed by its I 3 C NMR spectrum.

The aminoglycosicle antibiotics have provided a versatile backbone for the organic chemist to construct analogues with improved antimicrobial properties.2 While studies have concentrated on modifications of the 2-deoxystreptamine or on the aminosugar appended to its 0 - 4 position, fewer analogues have appeared with modified sugars on the 0-5 or 0-6 positions. Usually in nature, there is a furanoside at the 0-5 0022-3263/78/1943-3048$01.00,!0

position or a pyranoside at the 0 - 6 position, with the only reported 0-6 furanosides being the relatively inactive 6 - 0 (P-D-ribofuranosyl)paromamine,synthesized by Hanessian et al.3aand the corresponding neamine derivative synthesized by Suami et al.3bAs part of our program aimed at evaluating the antibacterial properties of pseudodisaccharides and pseudotrisaccharides derived from neamine, the (Y- and &(cis

0 1978 American Chemical Society

J . Org. Chem., Vol. 43, No. 15,1978

Aminoglycoside Antibiotics and trans)-aminofuranosylglycosides1 and 2, both isomers of kanamycin B,4 3, were prepared. Although 1 and 2 wlould

3049

favored because of anchimeric assistance in the active intermediate. On the other hand, a nonparticipating group, such as a nitrate ester or a benzyl ether, will favor the formation of a cis-glycoside. This occurs by kinetic control, whereby the alcohol displaces, with inversion, the statistically more prevalent (sterically less hindered) trans- (@)-halide.Based on these considerations, tribenzyl chloric!e 7b was chosen as the suitably protected glycosyl halide for the synthesis of 1 and 2. In addition, a-glycoside formation might be further enhanced by anchimeric a ~ s i s t a n c efrom ~ ~ ~the ~ 3-acetamido group present in 7b. In a first approach to 7b, the known diol 8,1° readily available from diacetone glucose, was benzylated to give crystalline

Hoq H 5

o&

OR

I

I OH

I OH

>“

10

CbH,CH,O

OMe

8,R=H 9, R = CH,C,H, kniiampcin B

4

both be interesting to evaluate, the a-(cis)-glycoside, 1, was expected to be the more desirable as it appears to be closer in structure, based on molecular models, to kanamycin B. The pseudodisaccharide starting material, 6, was obtained by acetylation and hydrolysis of the known diol 4.5Although

R,O* R,0

OR2 4.

R,:= H: RZ=

5, R ,

==

D x 3

COCH,: R.

6. R , =: COCH :

=

R. = H

OCH-C,H

7a, &-chloride b, 0-chloride

two hydroxyls are present in 6, adequate precedent exists in the literature3a96 to predict that a Koenigs-Knorr type condensation between it and a glycosyl chloride should give the desired 0 - 6 substituted pseudotrisaccharide. Although extensive efforts have been mounted to efficiently synthesize a-pyranosides,7 spurred by the presence of such a linkages in many natural products, only a limited number of publications (excluding nucleoside work) have appeared on the stereospecific preparation of ol-(l,Z-cis) furanoside^.^ The furanose system lacks the rigidity and anomeric effect of the pyranose ring, both useful in controlling glycosylations. In general, with a furanosyl halide, product control for glycosylation is determined by the type of substituent present at its 2 position. Thus, if the substituent is a participating group, such as an ester, trans- (6)-glycoside formation will be

11

diether 9. Acidic methanolysis of 9 under a variety of conditions yielded, instead of the desired methyl furanosides 10, a complex mixture of products from which a crystalline oxazoline dimethylacetal was isolated. The product was tentatively assigned the manno-structure 11 based on spectral data and mechanistic considerations. To avoid undesired oxazoline formation, the known azide 121° was used as the starting material and was successfully converted to the desired new crystalline chlorides 7a and 7b as outlined in Scheme I. In carrying out this series of reactions, nearly all of the azido intermediates were oils and required chromatography for purification. In the preparative work, intermediates were not purified until the crystalline mixture of acetamides 16a and 16b was obtained. Acetylation of free sugar 17 gave @-(trans)-acetate18, which, on solvolysis in ethereal hydrogen chloride, yielded the mixture of chlorides 7a and 7b. The higher melting, more polar a-chloride [NMR (CDC13) 6 6.12 (d, 1 H, J = 4 Hz)] could be selectively crystallized from the mixture. However, it quickly reverted to the original anomeric mixture [6 6.12 (d, -1/2 H, J = 4 Hz), 6.02 (s, -I/* H)] on standing in deuteriochloroform solution. Because of this rapid equilibration, the anomeric mixture was used in the pseudotrisaccharide work. The mixture of chlorides 7a and 7b was treated under anhydrous conditions for 5 days with an excess of diol 6 in dry methylene chloride containing pyridine and silver perchlorate. Chromatography of the crude product gave the desired crystalline a-glycoside, 19a, in 12%yield and the 6-glycoside, 19b, rSHCO.Me

19a, &-glycoside 19b, 0-glycoside

3050 J . Org. Chem., Vol. 43, No. 15, 1978

Sitrin, Cooper, and Weisbach assignments. In addition, Figure 1 shows the pH profile of 1 in the region containing its (2-4,C-5, and C-6 resonances (75-90 ppm). The fact that the two downfield peaks near 88 ppm both shift on protonation indicates that both glycosides are attached to carbons P to amino groups. Since one glycoside linkage (at C-4) was present in the starting neamine, the new linkage must be a t C-6 which is to the amino group on

Scheme I

c-1.

In vitro antibacterial testing of 1 against several Grampositive and Gram-negative organisms indicated it to have approximately 3% of the activity of kanamycin B whereas 2 had less than 1%the activity of kanamycin B. It is of interest that a-glycoside 1 is more active than &glycoside 2 although it would appear that the overall structure perturbation in going from a 3-aminopyranoside to a 3-aminofuranoside is too large for maintenance of the desired biological activity.

12, R = H 13, R = CH,C,H,

OR 14a+b, R = H; X = N,; CY- + 0-glycosides 1 5 a + b , R = CH,C,H,; X = N,; CY- + p-glycosides 1 6 a + b , R = CH,C,H,; X = NHAc; CY- + 8-glycohides

C,H CH 0 1

OCH,C,H 17

Ac,O

-

C b H i C H NHAc 2 0 W ~ c

HC1 ether

Pyr

7a

t 7b

I

OCH,C,H 18

in 1.5%yield. In the absence of pyridine, several other products and higher amounts of 19b were observed, resulting from anomerization and transglycosylation about the relatively acid-labile furanoside system. Although definitive structure assignments could not be made for 19a or 19b because of their complex NMR spectra, their structures were readily discerned from the simpler spectra of their respective deblocked final products 1 and 2. Both 1 and 2 display a doublet at 6 5.5 (J = 4 Hz) from the a-pyranosyl anomeric proton present in the neamine precursor. In addition, 1 has a doublet at 6 5.8 ( J = 5 Hz) indicative of a cis-(a)-anomeric furanosyl proton whereas 2 has a singlet a t 6 5.3 indicative of a trans-(P)-furanoside. As indicated in a previous paper in this series,laJ1 the location of attachment of the new sugar can be readily confirmed with the aid of pH profile nuclear magnetic resonance (I3C NMR) spectra. Table I shows the 13CNMRchemical shifts of synthetic a-glycoside 1, neamine (20), and a methylglycoside 21 a t various pHs with tentative resonance

,' ""-_

OH 20

OH 21

Experimental Section Column chromatography was carried out on J. T. Baker silica gel (60-200 mesh). Proton nuclear magnetic resonance spectra (NMR) were run on a Varian T-60 instrument using Me4Si as standard. Mass spectra were obtained from a Perkin-Elmer RMU-6 or a Varian CH5 instrument. Infrared spectra were run on a Perkin-Elmer Infracord Model 137. Carbon magnetic resonance (I3C NMR) spectra were recorded on a Varian CFT-20 instrument and calibrated with an internal (-5%) dioxane standard set at 67.4 ppm. Samples for I3C NMR spectra were decarbonated by passage through a short column of Amberlite IRA-400 (OH-) resin and lyophilized. All manipulations, including neutralizations with 38%DCl, were performed in a COz-free, nitrogen atmosphere. pDs were measured with pHydrion papers (Micro Essential Laboratories, Brooklyn, N.Y.) and are uncorrected for DzO. Melting points were taken in a Thomas-Hoover melting point apparatus and are uncorrected. Thin layer chromatography (TLC) was run on Uniplate precoated silica gel plates, 250 or 1000 pm (Analtech, Inc., Newark, Del.). Organic extracts were dried over anhydrous sodium sulfate. 5,6 -O,O'-Cyclohexylidene-2-deoxy-N,N'-bis(methoxycarbony1)-4-0 -[3,4-di- O-acetyl-2,6-dideoxy-2,6-bis[ (methoxycarbonyl)amino]-a-D-glucopyranosyl]streptamine ( 5 ) . T o a chilled solution of acetic anhydride (50 mL) in pyridine (200 mL) was added 5,6-0,0'-cyclohexylidene-2-deoxy-N,N'-bis(methoxycarbonyl)-4-0-[2,6-dideoxy-2,6-bis[(methoxycarbonyl)amino]-aD-glucopyranosyl]streptamine5(4) (100 g, 0.16 mol). The mixture was allowed to stand overnight at room temperature and concentrated in vacuo (0.1 mmHg). The residue was dissolved in methylene chloride (500 mL), washed with saturated aqueous sodium bicarbonate, dried, and concentrated in vacuo. The concentrate was triturated with ether and crude 5 (101 g, 88%) was collected. For analysis, a sample was chromatographed on silica gel with chloroform-methanol (97:3) to yield pure 5 as a white amorphous ~ ( e 1,CHC13);mass spectrum mle 718 (M+). Anal. solid: [ a I z 5+55.3O Calcd for C ~ O H ~ ~C,N50.14; ~ O H, ~ ~6.45; : N, 7.80. Found: C, 49.83; H, 6.47; N, 7.59. 2-Deoxy-N,N'-bis(methoxycarbonyl)-40-[3,4-di-O-acetyl-2,6-dideoxy-2,6-bis[ methoxycarbonyl)amino]-a-~-glucopyranosyllstreptamine (6). To a solution of cyclohexylidene diacetate 5 (100 g, 0.14 mol) in methanol (300 mL) was added 3 N aqueous HC1 (60 mL). The solution was allowed to stand for 6 h a t room temperature, neutralized with aqueous sodium bicarbonate, and concentrated in vacuo. The residue was dissolved in methylene chloride (500 mL), washed with water, dried, and concentrated in vacuo to give a tan oil which was dissolved in hot ethyl acetate (500 mL) and added dropwise to ether (2 L) with vigorous stirring to yield 6 (65.5 g, 73%) as an amorphous hygroscopic solid: [ a I z 5t77.8 ~ ( e 1, CHC13); MS mle 638 (M+). Anal. Calcd for C24H38N4016: C, 45.14; H, 6.00; N, 8.77. Found: C, 45.04; H, 6.29; N, 8.49. 3-Acetamido-5,6-di- 0-benzyl-3-deoxy- 1,2- 0,O'-isopropylidene-a-D-glucofuranose (9). Benzyl chloride (3.4 g, 27.0 mmol) was added dropwise with stirring to a mixture of 3-acetamido-3-deoxy1,2,-0,0'-isopropylidene-a-D-glucofuranose10 (8) (870 mg, 3.3 mmol), crushed potassium hydroxide (3.7 g), and crushed Drierite (3.7 g) in DMF (20 mL). The reaction mixture was stirred a t 65 "C for 2 h, allowed to cool, and filtered. The filtrate was concentrated in vacuo (0.1 mm) and the residual syrup was recrystallized from ether to give 9 (770 ~ ( e 1,CHC13); IR (CHC13) 1670 mg, 53%): mp 69-70 "C; [ m I z 5-27.4' and 1510 cm-I; NMR (CDC13) 6 5.8 (1H, d, J = 4 Hz), 1.75 ( 3 H, s), 1.50 (3 H , s), 1.30 (3 H , s); MS mle 426 (M - CH3)+. Anal. Calcd for

J . Org. Chem., Vol. 43, No. 15, 1978 3051

Aminoglycoside Antibiotics

Table -I. The C*3Chemical Shifts of 20,21, and 1 at pH (pD) 2 and 12 and Their Differences,= A Carbon atom 1 2 3 4 7

6 1' 2' 3' 4' 5'

6) 1" 2" 3" 4"

p H 1251.17 36.67 50.20 88.30 76.79 78.39 101.72 56.12 74.48 72.26 74.04 42.59

20 PH 2 50.64 29.02 49.43 78.16 75.95 73.25 96.47 54.44 69.98 71.55 69.03 41.11

s.

p H 12

21 PH 2

A

7.65 10.14 5.14 5.25 4.50 5.01 103.68 (78.32)

58.46 (77.96) 71.08 63.3:! 56.5b

,if' 6"

OMe

102.04 (75.38) 57.79 (75.16) 70.95 63.54 56.03

2.94 2.80

p H 12

1 PH 2

(50.54) 36.87 (50.10) (87.75) 75.40 (87.41) 101.37 56.18 74.47 72.28 74.08 42.55 104.50 (78.70) 57.92 (77.81) 70.86 64.21

(49.94) 28.33 (49.17) 78.24 74.97 82.27 96.79 54.38 70.04 71.49 69.03 41.02 103.03 76.09 56.66 74.79 70.94 6X38

s. 8.54 9.51 5.14 4.58 4.43 5.05 2.61 3.02

Chemical shifts in ppm from Me4Si. Shifts in parentheses were too close to be assigned unequivocally. procedure used to prepare 9 starting with 3-azido-3-deoxy-1,2O,O'-isopropylidene-a-D-glucofuranoselo (12) (17.3 g, 0.07 mol). Crude 13 was obtained as a syrup (29.5 g, 98%). For characterization, a sample was purified by preparative TLC with ethyl acetate-hexane (1:9) to yield 13 as an oil: [aI2'D -20.8' ( e 1, CHCl3); IR (film) 2110 cm-l (azide); MS r n l e 397 (M - N2)+, 382 (M - Nn - CH3)+. Anal. Calcd for C23H27N305: C, 64.93; H, 6.40; N, 9.88. Found: C, 65.21; H, 6.46; N, 9.83.

Methyl 3-Azido-5,6-di- 0-benzyl-3-deoxy-a- and -&D-glucofuranosides (14a and 14b). A solution of crude isopropylidene ketal 13 (29 g, 0.07 mol) in methanolic hydrogen chloride (770 mL, 1.4 % wlv) was allowed to stand overnight at room temperature and slowly

Figure 1. Change in chemical shifts (13C NMR spectra) of 1 with deuterium ion concentration (75-90 ppm only).

C Z ~ H ~ ~ N O VC, H 65.34; ~ O : H , 7.24; N, 3.05. Found: C, 65.18; H. 6.97; N, 2.95. 2-Methyl[ 5,6-di- O-benzyl-2,3-dideoxy-D-mannose(glucose)][3',2':4,5]-2-oxazoline Dimethyl Acetal ( 1 1). A solution of isopropylidene ketal 9 (1.9 g, 4.3 mmol) in methanolic hydrogen chloride (30 mL, 5% w/v) was warmed a t 60 "C for 3 h, quenched with aqueous sodium bicarbonate, and concentrated in vacuo. The residue was dissolved in chloroform, washed with aqueous sodium bicarbonate, dried, and evaporated under reduced pressure. Crystallization of the residue yielded oxazoline 11 (760 mg, 41%): mp 103-104.5 "C; +80.6" ( e 1,CHC1:3);IR (CHC13) 1670 cm-', no amide I1 band a t 1510 cm-'; NMR (CDC13)6 3.46 (6 H, s), 1.98 (3 H, d, J = 1Hz, five bond oxazoline coupling!, acetal proton hidden; MS mle 429 (M+), 414 (M - CH3)+, 354 (M - CH30CHOCH3)+,75 (CH30CHOCH3)+; mono-Me4Si derivative, MS m / e 501 (M 1 Me&)+, 486 (M + 1 Me4Si - CH3)+, 426 (M.t 1 Me4Si - CH30CHOCH3)+,75 (CH30CHOCH3)+. Anal. Calcd for Cz4H31NO~:C, 67.11; H, 7.27; N, 3.26. Found: C, 66.85; H. 7.40; N, 3.57. Acetylation of 11 with acetic anhydride in pyridine yielded a monoacetate: NMR (CDC1316 5.15 (1H, dd, J1 = 4 Hz, J z = 8 Hz), 3.47 (3 H, s), 3.43 (3 H, s), 2.09 (3 H, s), 2.00 (3 H, d, J = 1Hz), acetal proton hidden; IR (CDC13) 1740 and 1680 cm-', no NH or OH at 3000-4000 cm-', no amide I1 a t 1510 cm-l. 3-Azido-5,Gdi- 0-benzyl-3-deoxy-1,Z- 0,O'-isopropylidenea-D-glucofuranose (13). This compound was made by the same

+

poured into 2 L of ice water containing sodium bicarbonate (40 g). The methanol was removed by concentration in vacuo and the aqueous suspension was extracted with methylene chloride. The extracts were dried and concentrated in vacuo to give a mixture of 14a and 14b (26 g, 93%). For characterization a sample of the crude mixture was purified by preparative TLC (ethyl acetate-hexane, 1:4) to yield pure a-isomer 14a as an oil: R f 0.18 (ethyl acetate-hexane, 1:4); [ a ] 2 5+69.0' ~ (c 0.5, CHC13); IR (CHZClZ) 2110 cm-'; NMR (CDC13) 6 4.9 (1 H, d, J = 4 Hz), 3.4 (3 H, s); MS mle 370 (M - Na - HI+. Anal. Calcd for C21H25N305: C, 63.15; H, 6.31; N, 10.52. Found: C, 63.33; H, 6.56; N, 10.29. The (3 isomer, 14b, was also obtained from the preparative TLC as an oil: Rf 0.14; [Cf]25D -55.0' ( e 0.5, CHC13); IR (CHzC12)2110 cm-'; NMR (CDC13) 6 3.3 (3 H, s), anomeric proton hidden; MS mle 370. Anal. Found: C, 63.56; H, 6.55; N, 10.42.

Methyl 3-Azido-2,5,6-tri- 0-benzyl-3-deoxy-a- and -8-D-glucofuranosides (15a and 15b). The mixture of crude alcohols 14a and 14b (25 g, 0.062 mol) was benzylated in the same manner as used in the preparation of 9 to yield a mixture of crude 15a and 15b (33 g, 91%). For characterization the mixture was chromatographed using ethyl acetate-petroleum ether (4:96) to yield crystalline a-isomer 15a: mp t 48.1 ( c 0.5, CHC13); IR (CH2C12) 47.5-48" (ethanol-water); 2110 cm-l; MS mle 461 (M - Nz)+ and 460 (M - Nz - H)+. Anal. Calcd for C28H31N305: C, 68.69; H, 6.36; N, 8.58. Found: C, 68.42; H, 6.62; N, 8.34. The (3 isomer, 15b, was obtained as a thermally labile oil: CY]^^^ -28.8' ( c 0.5, CHC13); IR (film) 2110 cm-'; MS rnle 461, 460. Anal. Found: C, 69.05; H, 6.92; N, 8.30. Methyl 3-Acetamido-2,5,6-tri-O-benzyl-3-deoxy-aand -8glucofuranosides (16a and 16b). The crude mixture of tribenzyl azides 15a and 15b (11g, 0.022 mol) was stirred with lithium aluminum hydride (1 g) in ether (200 mL) for 2 h. The excess reagent was destroyed by sequential addition of water (1 mL), 10% aqueous sodium hydroxide (1 mL), and water (3 mL) and the suspension was filtered and concentrated in vacuo. The residue was stirred overnight in acetic anhydride (20 mL) and pyridine (20 mL), concentrated in vacuo, and crystallized from ether-petroleum ether to yield a mixture of 16a and 16b (7.1 g, 64%).

3052 J . Org. Chem., Vol. 43, No. 15, 1978

Sitrin, Cooper, and Weisbach

was concentrated in vacuo and the residue was refluxed for 4 h in For characterization, a sample was recrystallized from benzenewater (10 mL) containing barium hydroxide octahydrate (7.5 g). The cyclohexane to yield the more polar a isomer, 16a: mp 130-131 "C; solution was neutralized a t 100 "C with carbon dioxide gas and filtered R, 0.35 (ethyl acetate-hexane, 1:3); [ c I ] ~ ~ Dt63.3 (c 0.5, CHC1:3);IR and the filtrate was neutralized to pH 6 with 0.5 N sulfuric acid. The (CHC13) 1670 cm-'; NMR (CDC13) 6 4.76 (1 H, d, J = 4 Hz), 3.40 (3 sulfate was removed by filtration and the filtrate was charged H, s), 1.67 (3 H, s); MS m l e 505 (M+). Anal. Calcd for C ~ O H ~ ~C, N O ~ barium : onto a column of Amberlite IRC-50 ion exchange resin (20 mL) in the 71.27; H, 6.98; N, 2.77. Found: C, 70.98; H, 7.04; N, 2.83. ammonium cycle. The column was eluted with an ammonia gradient The less polar, (3 isomer, 16b, was isolated from the mother liquors: (0 to 0.5 M)and the fractions containing 1 were concentrated in vacuo, mp 104-105 "C (cyclohexane); R, 0.48; [O1Iz5D -36.7" ( c 0.5, CHCI,); neutralized to pH 6 with dilute sulfuric acid, and lyphilized to yield IR (CHC13) 1670 cm-'; YMR (CDC13) 6 3.3 (3 H, s), 1.77 (3 H, s), 4.85 1 as its sulfate salt (60 mg, 22%): [a]25D +88" (c 0.5, HzO); NMR (DzO) (1 H , s); MS mle 505 (R4+). Anal. Found: C, 70.97; H. 7.03; N, 2.85. 6 5.8 (1 H, d, J = 4 Hz), 5.5 (1 H, d, J = 5 Hz). Anal. Calcd for 3-Acetamido-2,5,6-tri- 0-benzyl-3-deoxy-D-glucofuranose ClsH37N~01~1.5H~S0~3HzO: C, 31.57; H, 6.77; N, 10.23; SO4,21.03. (17). A solution of the n~ethylglycosides16a and 16b (6 g, 0.012 mol) Found: C, 31.62; H. 6.20; N, 10.13;sod,21.21. in acetic acid (55 mL) containing 3 N hydrochloric acid (8 mL) was 2-Deoxy-4- 0-(2,6-diamino-2,6-dideoxy-a-~-glucopyranostirred at 40 "C for 3 h. The solution was cooled and poured into syl)-6- 0-(3-am~no-3-deoxy-~-D-g~ucofuranosyl)streptam~ne ice-water containing excess sodium bicarbonate. The semisolid pre(2). A sample of the protected @isomer,19b (597 mg, 0.5 mmol), was cipitate was collected and dissolved in methylene chloride and the deblocked in the same manner to yield 2 as its sulfate salt (258 mg, solution was washed w t h aqueous sodium bicarbonate, dried, and 55%):[ a I z 5+14.2" ~ (C 0.5, HzO); NMR (D20) 6 6.0 (1 H, d, J = 4 Hz), concentrated in vacuo. The residue was recrystallized from ben5.3 (1 H, s). Anal. Calcd for ClsH37N501r3H2SO~H20:C, 27.17, H, zene-cyclohexane to yield 17 (4.2 g, 73%): mp 104-105 "c;[@ID2' 5.70; N, 8.80; S042-, 36.21. Found: C, 26.85; H, 5.65; N, 8.38; S042-, -75.3' (C 0.5, CHC13); IVMR (CDC13 t D20) 6 5.45 (-'/* H, d, J = 4 37.61. Hz). 5.36 (-% H, s), 11.63 and 1.76 (3 H, 2s); MS mle 491 (M+). Anal. Calcd for C29H:I,jNOfj:C, 70.86; H, 6.77; N, 2.85. Found: C, 70.98; H , Acknowledgments. The authors wish to acknowledge the 7.02; N, 2.82. assistance of Dr. George Wellman, Mr. Thomas Ku, Ms. 3-Acetamido-1- O-acetyl-2,5,6-tri- O-benzyl-3-deoxy-fl-DJoanne Rosenbloom, and Mr. Norman Hall for preparation glucofuranose (18). A solution of 17 (4.2 g, 8.5 mmol) in acetic anhydride (10 mL) and pyridine (50 mL) was left at room temperature of some of the intermediates, Ms. Edith Reich for performing for 3 h and then concentration in vacuo. The residue was evaporated analyses and rotations, Mr. David Staiger for running the I3C several times with toluene and crystallized from benzene-cyclohexane NMR's, Mr. G. Roberts for running mass spectra, and Dr. Joe to yield 18 (3.2 g, 71%): mp 88-89 'C; [U]"D -49.7" (c 0.5, CHC13); Uri and Mr. Joe Guarini for biological data. NMR (CDC13) 6 6.15 (1 H, s), 1.84 (3 H, s), 1.80 (3 H, s); IR (CH2C12) 1750 and 1665 cm-I; M9 m / e 533 (M+).Anal. Calcd for C31H3:,NOi: Registry No.-1 sulfate salt, 66322-53-0;2 sulfate salt, 66290-55-9; C. 69.78; H, 6.61; N, 2.62. Found: C, 69.74; H, 6.76; N, 2.54. 4, 35017-15-3; 5, 57538-48-4; 6, 57538-46-2; 7a, 66290-72-0; 7b, 3-Acetamido-2,5,6-t,ri-0-benzyl-3-deoxy-a- and -fl-D-gluco66290-71-9;8,14166-54-2;9,66290-70-8; 11,66290-69-5; 11 monoafuranosyl Chlorides (7a and 7b). A solution of acetate 18 (3.2 g, 6 cetate derivative, 66290-68-4; 11 mono Me4Si derivative, 66290-67-3; mmol) in dry saturated ethereal hydrogen chloride (200 mL) con12,22169-71-7;13,66290-66-2;14a, 66290-65-1; 14b, 66290-64-0; 15a, taining acetyl chloride (20 mL) was stirred overnight a t room tem66290-63-9; 15b, 66290-62-8; 16a, 66290-61-7; 16b, 66290-60-6; 17, perature during which time a white precipitate formed and yedis66290-59-3; 18, 66290-58-2; 19a, 66322-52-9; 19b, 66290-57-1; 20, solved. This was concentrated in vacuo and evaporated repeatedly 3947-65-7;21,66290-56-0. with dry toluene to yield the anomeric mixture of chlorides 7a and 7b (3.0g, 98%):mp 114-115.5 "C; [.I2$ +114 +66" (c 0.5, benzene); References and Notes NMR (CDC13)6 6.12 (-J/2 H, d, J = 4 Hz, 01 anomer), 6.02 (-l/z H, s, $anomer), 1.74 (3 H, s); MS mle 473 (M - HCl)+;IR (CHzC12) 1670 (1) (a) Part I: R. D.Sitrin, D. J. Cooper, and J. A . Weisbach, J. Antibiot., 30, cm-'. Anal. Calcd for C19H32ClN05: C, 68.29; H, 6.32; C1,6.95;N, 2.75. 836 (1977); (b) Part 2: F. R. Pfeiffer, S.J. Schmidt, C. M. Kinzig. J. R. E. Found: C, 68.40; H, 6.3'7; C1, 6.71; N, 3.00. Hoover, and J. A. Weisbach, Carbohydr. Res., in press. (2) For reviews see S.Umezawa, Adv. Carbohydr. Chem., 30, 111 (1974); Crystallization from benzene-cyclohexane gave the higher moelting H. Umezawa, ibid., 30, 183 (1974); K. E. Price, J. C. Godfrey, and H.KaCY anomer, 7a [mp 120-121 "C; NMR (CDC13) d 6.12 (1 H, d, CJ = 4 waguchi, Adv. AppI. Microbiol., 18, 191 (1974); and R. Benveniste and J. Hz)], which equilibrated overnight in CDC13 solution to the original Davies. Antimicrob. Agents Chemother., 4, 402 (1973). anomeric mixture. (3) (a) T. Ogawa, T. Takamoto, and S. Hannessian, Tetrahedron Left., 4013 2-Deoxy -N-N'-bis( methoxycarbony1)-4- 0-[3,4-di- O-acet(1974); (b) T. Suami, S.Nishiyama, Y. Ishikawa, and S. Katsura. Carbohydr. Res., 52, 187 (1976). yl-2,6-dideoxy-2,6-bis[ (methoxycarbonyl)amino]-a-D-gluco(4) T. Ito, M. Nishio, and H. Ogawa, J. Antibiot., Ser. A, 17, 189 (1964). pyranosyll-6- 0-(3-acetamid0-2,5,6-tri-O-benzyl-3-deoxy-a(5) T. Jikihara, T. Tsuchiya, S. Umezawa, and H. Umezawa, Bull. Chem. SOC. and -8-D-g1ucofuranosyl)streptamines(19a and 19b). A mixture Jpn., 46, 3507 (1973). of chlorides 7a and 7b (1.5 g, 2.9 mmol), the protected neamine 6 (3.6 (6) A. Hasegawa, N. Kurihara, D. Nishimura, and M. Nakajima, Agric. Biol. g, 6.6 mmol), finely ground Drierite (10 g), and pyridine (0.25 niL) in Chem., 32, 1130-1 134 (1968); S.Umezawa, S1 Koto, K. Tatsuta, H. Hineno. Y. Nishimura. and T. Tsumura. Bull. Chem. SOC.Jon.. 42.537-541 dry methylene chloride 140 mL) was stirred with exclusion of moisture (1969); R. U. Lemieux, T. Nagabhushan, K. Clemetson, a i d L. Tucker, Can. for 2 h. and then anhydrous silver perchlorate (890 mg) was added. J. Chem., 51, 53 (1973). Stirring was continued in the dark for 5 days, after which the reaction (7) For reviews, see G. Wolff and 0. Rohle, Angew Chem., Int. Ed. Engl., 13, was quenched with aqueous sodium bicarbonate and filtered. The 157 (1974); C. Schuerch, Acc. Chem. Res., 6, 184 (1973); and R . U. filtrate was extracted several times with methylene chloride and the Lemieux, K. B. Hendriks, R. V . Stick, and K. James, J. Am. Cbem. SOC., 97, 4056 (1975). combined organic phases were dried and concentrated in vacuo. The (8) (a) R. K. Ness and H. G. Fletcher, J. Org. Chem., 22, 1465 (1957). J. Am. residue (2.53 g) was chromatographed (chloroform-methanol, 98:2) Chem. Soc., 80, 2007 (1958); (b) P. A. J. Gorin, Can. J. Chem., 40, 275 to yield the more mobile 6-isomer 19b (55 mg, 1.5%): [a]2 5 -2" ~ (c (1962); (c) C. P. J. Glaudemons and H.G. Fletcher, J. Org. Cbem., 29, 3286 0.5, CHC13);IR ( CHCIZ3) 1730 and 1670 cm-'; R, (chloroform-meth(1964); J. Am. Chem. SOC.,87, 2456, 4636 (1965). anol, 982) 0.54. .4nal. Calcd for C53HfigN5021:C, 57.40; H, 6.00; N , 6.31. (9) D. Nishimura, A. Hasegawa, and M. Nakajima, Agr. Biol. Chem., 36, 1767 (1972). Found: C, 57.15; H, 6.21; N , 6.59. (10) W. M. zu Reckendorf, Angew. Chem., Int. Ed. Engl., 5 , 967 (1966); Cbem. Further elution yielded the more polar, crystalline a-isomcr 19a Ber., 101,3802 (1968); D. T. Williamsand J. K. N. Jones, Can. J. Chem., (407 mg, 12%based on '7a and 7b): mp 201-203 "C (toluene-metha45,7 (1967); J. S.Brimacombe. J. G. H. Bryan. A. Husain, M. Stacey. and no]); [CY]25D +63.6 (c 0.5. CHC13);IR (CHC13) 1730 and 1670 cm-'; R f M. S.Toiley, Carbohydr. Res., 3, 318 (1967). 0.39. Anal. Found: C, 57.14; H, 6.10; N, 6.18. (11) P. W. K. Woo and R. D. Westland, Carbohydr. Res., 31, 27 (1973); J. B. Morton. R. C. Long, P. J. L. Daniels, R. W. Tkach, and J. H. Goldstein, J. Am. 2-Deoxy-4- 0-(2,6-diamino-2,6-dideoxy-a-D-glucopyranoCbem. SOC.,95,7464 (1973); S. Omoto, S. Inouyl, M. Kojima. and T. Niida, sy1)-6- 0-(3-am~no-3-deoxy-a-D-glucofuranosy~)streptam~ne J. Antibiot., 26, 717 (1973); K. F. Koch, J. A. Rhoades, E. W. Hagman, and (1). The recrystallized protected a-pseudotrisaccharide 19a (465 mg, E. Wenkert, J. Am. Cbem. Soc., 96,3300 (1974); T. L. Nagabhushan and 0.4 mmol) in ethanol (20 mL) was hydrogenated a t 60 psi over 10% P. J. L. Daniels, TetrahedronLeft., 747 (1975); T. L. Nagabhushan. W. N. palladium on charcoal (500 mg) for 16 h, after which the catalyst was Turner, P. J. L. Daniels, and J. B. Morton, J. Org. Chem., 40, 2830 (1975). removed by filtration and washed with excess ethanol. The filtrate

-